By DigiBoss Editorial Team · Last updated: October 16, 2026 · 9 min read

Why Database Quality Decides Campaign Results

Two files can have the same number of rows and give completely different results. One contains current records that match your audience. The other contains old, duplicated or irrelevant entries. The first lets your telecallers talk to real prospects. The second burns hours on wrong numbers and unreachable people.

Quality problems also cost more than time. High email bounce rates harm your sender reputation, and many complaints can get your numbers or accounts restricted. Checking quality before you buy is the cheapest risk control a marketer has.

This audit works for any file, whether you buy from DigiBoss or from another seller. If you are still deciding what type of file to buy, start with our complete guide to lead databases in India.

The Six Dimensions of Data Quality

Step 1: Ask Where the Data Comes From

Ask the vendor three plain questions and note the answers. How was the data collected? Did people agree to be contacted? How often is it refreshed? You are not looking for secret details. You are looking for a clear, confident explanation.

Vague replies such as “it is all verified” or “fresh data” are not answers. Ask what verified means. Verification might mean a phone call, an OTP check or a syntax check on email addresses. Each has a different value. Keep the reply with your purchase records, because data protection rules expect you to know your sources. See our guide to the DPDP Act for marketers.

Step 2: Request a Sample and Read the Fields

Ask for a masked or small sample that shows the same columns as the full file. For student data, check that fields such as state, stream and board match what you need. For business data, check company, designation, industry and city. Our homepage example shows a typical masked layout, and each product page lists fields. Sample sizes and fields differ by product, so ask the seller for details.

Never judge a file by the sample alone, because samples can be hand-picked. Treat the sample as a way to test the structure and to run the next checks.

Step 3: Check Completeness

Open the sample in Excel and count how many rows have every field you need. A file with 100 percent of names but only half of the states is hard to segment by region.

• Select the column and look at the count in the Excel status bar, or use =COUNTA(B2:B501) to count filled cells.
• Use =COUNTBLANK(B2:B501) to count empty cells in a column.
• Calculate the filled share as =COUNTA(B2:B501)/ROWS(B2:B501).

Decide your pass mark before you start. For example, you may require nearly all rows to have a valid mobile number and a state, and treat other fields as optional.

Step 4: Test Mobile Numbers and Emails for Format

Indian mobile numbers are ten digits and normally start with 6, 7, 8 or 9. Numbers stored with a +91 prefix or spaces need cleaning first. Use these simple tests on a clean copy of the sample:

• Length test: =LEN(C2)=10 returns TRUE for ten characters.
• First digit test: =OR(LEFT(C2,1)="6",LEFT(C2,1)="7",LEFT(C2,1)="8",LEFT(C2,1)="9").
• Repeating pattern test: sort the column and look for numbers like 9999999999 or 1234567890.
• Email syntax test: =AND(ISNUMBER(SEARCH("@",D2)),ISNUMBER(SEARCH(".",D2))) flags rows missing basic parts.

These tests catch format problems, not whether a person still uses the number. For that, you need the pilot in Step 7. Our guide to cleaning a lead list in Excel shows how to fix many issues before you test.

Step 5: Find Duplicates

Duplicates lead to the same person receiving repeated calls or emails, which creates complaints. Check them in two ways.

1. Use Data, then Remove Duplicates on a copy of the file, selecting the mobile number column. Note how many rows disappear.
2. Or, in a helper column, use =COUNTIF($C$2:$C$501,C2) and filter for values above 1.

Also compare the new file with your existing CRM. If a large share of records are already in your system, the true value of the purchase is lower than the row count suggests.

Step 6: Check Freshness and Fit

Ask when the data was last updated and whether the audience still matches your target. A list of Class 12 students from two exam cycles ago is now a list of college students or working adults. For exam-driven segments, timing matters. Match your purchase to your campaign calendar using our education marketing calendar.

Fit also includes geography. If you teach in Patna, a file that is mostly from another region has limited value even when it is accurate.

Step 7: Run a Small Pilot

A pilot is the only way to learn how the whole file behaves. Take a random slice of about 5 to 10 percent and contact it using your normal method, within the legal limits in TRAI telemarketing rules and the data law.

Set your own pass marks from past campaigns. If you have none, treat the pilot as your baseline and compare the next file against it.

Step 8: Read the Policies and Support Terms

Digital files usually have strict refund terms because they cannot be returned. Read the refund, privacy and terms pages first, and ask what happens if the file does not match its description. Confirm that a real person can answer by call or WhatsApp. If you plan to share the file with agencies or telecallers, confirm what the licence allows.

Step 9: Score the Vendor and Decide

Add the scores. Set your threshold before you start, for example a minimum of 11 out of 14. If the file fails, ask the vendor to fix specific issues or move on. Do not let a discount push you past a failed check.

Red Flags in Lead Data Offers

• Extremely low prices for millions of records with no source explanation.
• No sample details, no refund terms and no contact person.
• Claims of 100 percent accuracy. No large list is perfectly accurate.
• Pressure to buy immediately.
• Files that arrive as locked or unreadable formats.
• Free downloads from unknown sites. See the risks of free student number downloads.

Worked Example: Auditing a 500-Row Sample

Imagine you receive a 500-row sample of Class 12 student data. Here is how you could audit it in about twenty minutes. The numbers below are illustrative.

A clear record like this lets you negotiate with facts. You can ask for a replacement of weak records or a lower price, or you can decide to walk away.

Copy-and-Paste Questions for Any Vendor

1. Where does the data come from and when was it last updated?
2. What does verified mean for this file: calls, OTP, syntax checks or something else?
3. Which fields are included and how complete are they?
4. Does the file include under-18 students, and how should I handle them?
5. Can I get a sample with the same columns as the full file?
6. What is the refund or replacement policy for digital files?
7. How can people ask to be removed from the data?
8. Who can I contact after the sale?

Paste these into an email or WhatsApp message and keep the replies. Clear answers are a good sign. Silence or confusion is useful information too.

When to Walk Away

• The vendor will not explain the source in any detail.
• The sample differs sharply from the structure or quality promised.
• More than a small share of mobile numbers fail format checks.
• Duplicates are common and the vendor refuses to remove them.
• The refund and support terms are missing or unclear.
• The price is far below the market with a claim of perfect accuracy.

Walking away costs nothing. Buying a bad file costs money, time and reputation. For a broader view of buying decisions, see the lead database guide.

Template: Data Quality Log

Keep a simple log for every file you evaluate or buy. A one-page record helps you compare vendors and defend your decisions later.

Over time, this log shows which vendors and segments perform best for your business. That knowledge is worth more than any single file.

Frequently Asked Questions

How can I check if a database is genuine?

Ask about its source and update date, request a sample, test it in Excel for completeness, format and duplicates, then run a small pilot and review answer, bounce and opt-out rates.

How many mistakes are acceptable in a purchased list?

No list is perfectly accurate. Set your own tolerance based on past campaigns and the cost of each contact, and use a pilot to measure the real rate before scaling.

What Excel formula finds duplicate mobile numbers?

In a helper column use COUNTIF on the mobile column, for example =COUNTIF($C$2:$C$501,C2), then filter for values above 1. You can also use Data, Remove Duplicates on a copy.

How do I check if a mobile number format is valid in India?

A standard Indian mobile number has ten digits and usually starts with 6, 7, 8 or 9. Remove spaces and the +91 prefix first, then test length and the first digit.

Should I test a database before buying?

Yes. Request a sample, audit it in Excel and, if possible, run a small pilot on a slice of the file before committing to a large purchase or a full campaign.

How often should I re-audit my database?

Review it before every major campaign and at least once a year. Student and business contacts change quickly, so older files lose value.

Sources and Further Reading

• Microsoft Support: Remove duplicate values – official Excel help for removing duplicates and using functions
• TRAI (official site) – rules on commercial calls and SMS
• MeitY (Government of India) – data protection law and rules